This article was downloaded by: [Christian Thierfelder] On: 25 January 2013, At: 21:48 Publisher: Taylor & Francis Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK International Journal of Agricultural Sustainability Publication details, including instructions for authors and subscription information: http://www.tandfonline.com/loi/tags20 Integration of conservation agriculture in smallholder farming systems of southern Africa: identification of key entry points Christian Thierfelder a , Talkmore Mombeyarara b , Nelson Mango b & Leonard Rusinamhodzi b a CIMMYT, P.O. Box MP 163 Mt Pleasant, Harare, Zimbabwe b CIAT, P.O. Box MP228 Mt Pleasant, Harare, Zimbabwe Version of record first published: 25 Jan 2013. To cite this article: Christian Thierfelder , Talkmore Mombeyarara , Nelson Mango & Leonard Rusinamhodzi (2013): Integration of conservation agriculture in smallholder farming systems of southern Africa: identification of key entry points, International Journal of Agricultural Sustainability, DOI:10.1080/14735903.2013.764222 To link to this article: http://dx.doi.org/10.1080/14735903.2013.764222 PLEASE SCROLL DOWN FOR ARTICLE Full terms and conditions of use: http://www.tandfonline.com/page/terms-and-conditions This article may be used for research, teaching, and private study purposes. Any substantial or systematic reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form to anyone is expressly forbidden. The publisher does not give any warranty express or implied or make any representation that the contents will be complete or accurate or up to date. The accuracy of any instructions, formulae, and drug doses should be independently verified with primary sources. The publisher shall not be liable for any loss, actions, claims, proceedings, demand, or costs or damages whatsoever or howsoever caused arising directly or indirectly in connection with or arising out of the use of this material.
Transcript
This article was downloaded by: [Christian Thierfelder]On: 25 January 2013, At: 21:48Publisher: Taylor & FrancisInforma Ltd Registered in England and Wales Registered Number: 1072954 Registered office:Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK
International Journal of AgriculturalSustainabilityPublication details, including instructions for authors and subscriptioninformation:http://www.tandfonline.com/loi/tags20
Integration of conservation agriculture insmallholder farming systems of southernAfrica: identification of key entry pointsChristian Thierfelder a , Talkmore Mombeyarara b , Nelson Mango b &Leonard Rusinamhodzi ba CIMMYT, P.O. Box MP 163 Mt Pleasant, Harare, Zimbabweb CIAT, P.O. Box MP228 Mt Pleasant, Harare, ZimbabweVersion of record first published: 25 Jan 2013.
To cite this article: Christian Thierfelder , Talkmore Mombeyarara , Nelson Mango & LeonardRusinamhodzi (2013): Integration of conservation agriculture in smallholder farming systems ofsouthern Africa: identification of key entry points, International Journal of Agricultural Sustainability,DOI:10.1080/14735903.2013.764222
To link to this article: http://dx.doi.org/10.1080/14735903.2013.764222
PLEASE SCROLL DOWN FOR ARTICLE
Full terms and conditions of use: http://www.tandfonline.com/page/terms-and-conditions
This article may be used for research, teaching, and private study purposes. Any substantialor systematic reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, ordistribution in any form to anyone is expressly forbidden.
The publisher does not give any warranty express or implied or make any representation that thecontents will be complete or accurate or up to date. The accuracy of any instructions, formulae,and drug doses should be independently verified with primary sources. The publisher shall notbe liable for any loss, actions, claims, proceedings, demand, or costs or damages whatsoever orhowsoever caused arising directly or indirectly in connection with or arising out of the use of thismaterial.
A component-omission experiment based on the principle of conservation agriculture (CA)was established on smallholder farms for three seasons in Murehwa and Hwedza districts,Zimbabwe; Barue district in Mozambique; Balaka district and Chitedze Research Station inMalawi, and Monze district in Zambia to identify strategies for improving crop productivityand livelihoods for smallholder farmers. The objective of the experiment was to evaluate theeffect of tillage, residue retention, fertiliser application and weed control on maize yield. Inaddition, the study analysed possible combinations of these factors that could provide asustainable entry point for intensification through CA. Results showed that fertilisation hadthe strongest effect on crop yield in both tillage systems; adequate fertilisation is thereforekey to success in CA. Retention of crop harvest residues increased yield in no-tillagesystems; no-tillage without residues depressed yield by 50% when compared with yields ofconventional tillage. A step-wise integration of CA into the smallholder farming systems isproposed as a possible strategy to avoid new constraints on smallholder farms. If resourcesare limiting, farmers may apply all principles on small areas to overcome the initial demandin resources (labour, fertiliser and residues), and once productivity is raised, they can expand.
Smallholder agriculture in southern Africa is constrained in many areas by low soil fertility, fre-quent droughts or excessive water, water run-off and soil erosion, inappropriate and often dys-functional input–output markets and weak extension systems. As a result, many ruralhouseholds in the region are malnourished, cannot improve their livelihoods and are food insecure(Dixon et al. 2001, Vanlauwe et al. 2010). Yields are declining with the current agriculturesystems (Figure 1), and land pressure prevents farmers from shifting to more fertile virginland, and therefore more inputs are required to stem soil fertility decline (Smaling et al. 1997).
Maintenance of soil organic matter over time is important for restoration of soil fertility andsustainability of agriculture systems (Bayer et al. 2000, Bessam and Mrabet 2003, Wall 2007).Unless farmers can maintain or increase the level of organic matter in their soils, the agriculturalproduction systems cannot be environmentally sustainable, and lead to physical, chemical andbiological soil degradation in the long term (Thierfelder and Wall 2011). Among the sustainableoptions to restore soil fertility and improve crop productivity is conservation agriculture (CA)(Wall 2007, Kassam et al. 2009, Thierfelder and Wall 2009). CA is based on minimum soil dis-turbance, surface crop residue retention (mulching) from previous crops or cover crops, and
diversified crop rotations or associations (FAO 2002). CAwas developed in the America and Aus-tralia mostly on large-scale farms and has shown tremendous success in these environments (Bol-liger et al. 2006). However, the feasibility of CA for smallholder farmers in southern Africa isconstrained by biophysical and socio-economic challenges (Giller et al. 2009). CA requiresincreased knowledge of the system and adaptation to the site and farmer circumstances (Wall2007, Erenstein et al. 2012). A special set of cultural practices that are different from traditionalplough-based systems need to be acquired (Lal et al. 2000). Adaptation to new seeding techniquesand fertilisation strategies, new residue management and weed control options, different harvestand crop management procedures is important for the successful implementation of CA (Thier-felder and Wall 2011). Providing evidence that crop production is possible without tilling thesoil is the most important stage in encouraging a shift in mindset among farmers so they canmove away from the traditional mouldboard plough (Wall 2007).
The integration of CA into smallholder farming systems has to address the prevailing con-straints without creating new complex ones. Unlike a new seed variety or a new type of fertiliser,CA involves many simultaneous changes to the farming system, which can lead to complexity(Thierfelder and Wall 2011). However, higher and stable yields are often obtained when all com-ponents of CA are implemented. The benefits of CA cannot be realised when soils are depleted ofplant nutrients; the soils will not support the production of enough biomass needed for cropresidue retention. The obvious and widely accepted strategy to address these constraints wouldbe the use of mineral fertilisers. However, the use of fertiliser among smallholder farmers insouthern Africa remains low due to limited market access and high prohibitive cost – i.e. fertilisercosts are between two and six times the cost in America, Europe and Asia (Smaling et al. 1997,Sanchez 2002). Some soils do not respond to added chemical fertiliser due to extremely smallconcentrations of soil organic matter (Tittonell et al. 2007). Depending on the farming system,
Figure 1. Average maize yield in Zimbabwe (in t ha– 1) from 1970–2004 (CSO, 1987; CSO, 1984–1989;FAOSTAT, 2004).
an integrated approach to restore soil nutrients should be pursued along with the practice of CA.Among the possible solutions to overcome soil nutrient limitations are inorganic fertilisers, grainlegumes, animal manures, compost, integrated nutrient management and agroforestry (Mafon-goya et al. 2006, Gilbert 2012).
Research results from Latin America show the importance of crop residues in managing CAsystems (Wall 1999, Govaerts et al. 2006, Verhulst et al. 2010) and most of the benefits in CA arelinked to crop residue retention (Erenstein 2002). In southern Africa, smallholder farmers managemixed crop–livestock systems where animals are used for draught power, produce manure forcrop production, contribute to income and risk reduction and document the status of wealth ina community (Valbuena et al. 2012). Competing uses of crop residues for soil surface coverand for livestock feed create conflicts and there are strong trade-offs for their efficient allocation(Mueller et al. 2001, Thierfelder and Wall 2012).
Another critical factor in the integration of CA is provision of sufficient knowledge to farmersand extension agents on the effects of each component of the CA system on crop productivity.Farmers understand the effect of fertilisation on crop yield, but the effects of reduced tillageand/or residue retention and the combination of these are relatively unknown and little under-stood. The contributions of each component of the system are also difficult to separate as mostof the benefits of the CA systems arise when several components are integrated with each other.
The objective of this study was therefore to analyse the effect of CA component-omission oncrop yield in multi-locational trials in Malawi, Mozambique, Zambia and Zimbabwe. The factorstested were tillage, residue retention, fertilisation and weed control. However, one of the threeprinciples of CA – crop rotations and/or associations – was not included. The analysis wasalso intended to identify a sustainable pathway for the integration of CA into the smallholderfarming systems of southern Africa.
2. Materials and methods
2.1. Sites description
The trials were carried out in Balaka District (14.99 S; 34.97 E, altitude: 622 m a.s.l.), Malawi infive intervention villages (Ntonya, Zammimba, Njereka, Chifodya and Chimkwezule) and at theChitedze Research Station (13.97 S; 33.65 E, altitude 1,144 m.a.s.l), Malawi; in Barue District,Mozambique (18.11 S; 33.19 E, altitude: 590 m a.s.l.) in four intervention villages (Mussianharu,Munene, Mvilamiti, Nyamuka), five villages (Wagoneka, Samunderu Nyamutsika Chidhora Nhu-karume) in Hwedza (18.37 S; 31.35 E altitude 1,427 m.a.s.l) and four intervention villages(Bruces, Springdale, Kournine, Twin Rivers) in Murehwa Districts (17.74 S: 31.57 E altitude1,280 m.a.s.l), Zimbabwe and the Malende Agriculture Camp in Monze District (16.24 S;27.44 E; altitude: 1,103 m a.s.l.), Zambia (Figure 2).
Harvest data were available for three cropping seasons (2008/2009–2010/2011) in Balakaand Barue, two cropping seasons (2009/2010 and 2010/2011) in Chitedze and Hwedza andeach one cropping season in Murehwa (2009/2010) and Monze (2010/2011).
All target areas were dominated by maize (Zea mays L) production with the highest intensityof this crop being found in Malawi. The sites in Zimbabwe and Zambia are characterised byhigher integration between crop and livestock components where crop residues are fed to live-stock and manure used for crop production. In Mozambique and Malawi, the level of integrationdiminishes; most of the sites in Malawi are crop-based farming systems. Rotations with othercrops were not widespread in Malawi due to land constraints and relatively small land holdings(World Bank 2007). Groundnuts (Arachis hypogea) and tobacco (Nicotiana tabacum) and some-times cassava (Manihot esculenta) and tomatoes (Lycopersicon esculentum) are common
International Journal of Agricultural Sustainability 3
rotational crops in Balaka and around Chitedze. In Mozambique farmers grow sorghum (Sorghumbicolour L.) besides maize or intercrop maize with pigeonpea (Cajanus cajan L.). Common beans(Phaseolus vulgaris) and sunflower (Helianthus annuus) were grown as cash crops. In Zim-babwe, maize is the main food crop followed by sorghum and groundnuts. Farmers growbeans, sweet potatoes (Ipomoea batatas), finger millet (Eleusine coracana) and sunflower inrotation or on homestead fields.
The experiments in Balaka, Malawi were established on soils with a sandy-to-sandy loamsurface soil texture and average rainfall of about 855 mm per season; at Chitedze ResearchStation, soils have mainly sandy clay loam texture and the average rainfall is around 960 mm.The sites in Mozambique were mainly on clay loams and the average rainfall of about1,000 mm per season. The sites in Murehwa, Zimbabwe had mainly sandy loam soils and anaverage rainfall of 850 mm per season. In Hwedza, soils were sandy and the sites received thelowest average rainfall of 600 mm per season. At Malende, Zambia the soils are sandy clayloams with the average rainfall of 748 mm per season.
Figure 2. Map of the regional trials in Balaka District and Chitedze Research Station of Malawi, BarweDistrict of Mozambique, Hwedza and Murehwa District of Zimbabwe and Monze District of Zambia.
4 C. Thierfelder et al.
Dow
nloa
ded
by [
Chr
istia
n T
hier
feld
er]
at 2
1:48
25
Janu
ary
2013
2.2. Experimental design
The experiment called ‘step trials’ was established in all sites in 2008. It was designed around aseries of increasing management steps from a very basic conventional system to a high-inputintensive CA system (Table 1). All sites had no previous history of no-tillage and were establishedon farmers’ fields previously under conventional tillage practices. The treatments consisted of (1)a farmer’s practice with conventional tillage and no fertiliser (CP); (2) conventional tillage andmineral fertiliser (CP+F); (3) a no-tillage system with no fertiliser and no residue retention(NT); (4) no-tillage, with no fertiliser but residue retention (NT+R); (5) no-tillage treatmentwith no residue retention but fertiliser (NT+F); (6) no-tillage with fertiliser and residue retention(NT+F+R); and (7) a no-tillage treatment with fertiliser, residue retention and chemical herbi-cide use (NT+F+R+H). Treatment 3 was not applied on all the sites in the 2008/2009 croppingseason. The treatments are replicated four times at each location and randomized in a completelyrandomized block design.
The seven common treatments were slightly different at each site due to different land prep-aration techniques and local fertiliser application rates. In Malawi the hand hoe was used to makeridges and furrows in the conventional systems and residues were removed. The CA systems wereplanted with a pointed stick (dibble stick) and residues applied at a rate of 2.5–3 t ha– 1 if treat-ments had residues. The population density followed the Sassakawa Global 2000 recommen-dation of 53,000 plants per hectare, planted in 75 cm rows and 25 cm in-row spacing (Itoet al. 2007). Plots were fertilised at 69N:21P2O5:4S if treatments had fertiliser, supplied in abasal and top dressing at about 4 weeks after planting. Weed control in treatment 7 was amixture of 2.5 l ha– 1 glyphosate (N-(phosphono-methyl)glycine) and 6 l ha– 1 of Bulletw(25.4% Alachlor (2-chloro-N-(2,6-diethylphenyl)-N-(methoxymethyl) acetamide) and 14.5%atrazine (2-chloro-4-ethylamino-6-isopropylamino-1,3,5-triazine) which was applied as a pre-emergence herbicide after planting. All treatments were additionally weeded with a hand hoewhen weeds were 10 cm high or 10 cm in circumference. Weeding on all no-tillage treatmentswas shallow to avoid soil disturbance as much as possible.
In Mozambique, land preparation was manual using hand hoes in the conventional practices.Residues in treatments 1 and 2 were burned in situ. In the CA treatment with residues they wereretained and a manual jab-planter was used for planting. Fertiliser was applied at a rate of58N:24P2O5:12 K2O as a basal dressing and one topdressing at 5 weeks after planting if treat-ments had fertiliser. Weed control was done manually with hand hoes except for treatment 7 inwhich glyphosate was applied at a rate of 2.5 l ha– 1 followed by manual weeding.
In Hwedza and Murehwa, Zimbabwe, treatments 1 and 2 were seeded into the soil prepared bythe mouldboard plough and into planting basins in treatments 3–7. All residues were removed inthe conventional tillage treatments whereas they were retained in situ on the CA plots if residues
Table 1. Trial design of the multi-locational ‘step-trial’ in Malawi, Mozambique, Zambia and Zimbabwe.
1 Control, +T, 2F 3 W W W2 Control, +T +F 3 3 W W3 No-till, 2F, 2R W W W W4 No-till, 2F, +R W W 3 W5 No-till, +F, 2R W 3 W W6 No-till, +F, +R W 3 3 W7 No-till, +F, +R, +H W 3 3 3
Notes: Treatment 3 was tested in the 2009/2010 season only.3 – yes; W – no; T – tillage; F – fertiliser; R – residues; H – herbicides.
International Journal of Agricultural Sustainability 5
were used in the treatments. Fertiliser was applied in fertilised treatments at a rate of81N:24P2O5:12K2O as a basal and top dressing. Weed control in treatment 7 was achievedwith glyphosate at a rate of 2.5 l ha– 1 and manual weeding. All other treatments were weededwith hand hoes four times during each season to reduce weed pressure on crop productivity.
The site in Malende, Zambia followed the same land preparation and management practices asin Hwedza and Murehwa but fertilization was slightly higher at a rate of 109N:34P2O5:17K2O.For weed control, only glyphosate at a rate of 2.5 l ha– 1 and manual weeding was applied. Allother treatments were weeded with hand hoes as necessary.
2.3. Yield measurement
At physiological maturity, maize was harvested from four rows by 5 m from each plot. Theharvest area of the net plots was used to extrapolate yields to a hectare basis. A sub-sample ofcobs per plot was dried and shelled to calculate grain yield at 12.5% moisture. All maizestalks were weighed at harvest without harvest produce (grain or cobs) and recorded asbiomass; one biomass subsample per plot was air dried for at least 4 weeks before final dryweights were taken and biomass was calculated to an area basis.
Harvest data were available for three cropping seasons (2008/2009–2010/2011) in Balakaand Barue, two cropping seasons (2009/2010 and 2010/2011) in Chitedze and Hwedza andeach one cropping season in Murehwa (2009/2010) and Monze (2010/2011).
2.4. Statistics
Statistical analyses were carried out using Statistix 9 for Windows (Statistix 2008). Yield datawere tested for normality and subjected to an analysis of variance using completely randomizedblock design. Where the F-test was significant a least significant difference (LSD) test was used atp , 0.05 for mean separation. Binary recursive portioning was used for constructing classifi-cation trees by splitting the data into homogeneous binary subgroups or nodes based on thefactors tested in this experiment. Node splitting was based on the Gini splitting rule:
Gini(t) = 1 − Sip2i
where pi is the probability of class i in t (Apte and Weiss 1997). The Gini splitting procedurefinds the largest homogeneous category within the dataset and separates it from the remainder ofthe data; subsequent nodes are then identified the same way until further divisions are not possible(Buntine 1992).
3. Results
3.1. Crop yield trends
Results showed that yield benefits were not universal (Table 2); in Balaka, yield ranges were777–2,675 kg ha– 1 in the 2008/2009 season, 1,774–5,180 kg ha– 1 in the 2009/2010 seasonand 1,281–3,269 kg ha– 1 in the 2010/2011 season. At the Chitedze Research Station yieldranged from 1,667 to 4,321 kg ha– 1 in the 2009/2012 season and 1,942–5,524 kg ha– 1 in the2010/2011 season. In Barue, yield ranges were 894–1,409 kg ha– 1 in the 2008/2009 season,1,974–4,099 kg ha– 1 in the 2009/2010 season and 1,272–4,168 kg ha– 1 in the 2010/2011season. In Hwedza, yields ranged between 866 and 2,486 kg ha– 1 in the 2009/2010 seasonand then 307–975 kg ha– 1 in the 2010/2011 season. In Murehwa, yields ranged between
1,223 and 3,114 kg ha– 1 in the 2009/2010 season. In 2010/2011, the site was hit by drought andno crop was harvested. In Monze, yield ranges were 3,002–5,904 kg ha– 1 in 2010/2011.
Across sites, yields were depressed in the 2008/2009 season but increased significantly in thefollowing two seasons. In the absence of crop residue cover, conventional tillage resulted in
Table 2. Crop yield as affected by tillage, mulch, fertiliser and weed control across six districts and threeseasons in southern Africa.
Site Treatment
Year
2008/2009 2009/2010 2010/2011
Balaka, Malawi Conventional tillage 777 2,032 1,281Conventional tillage + F 2,203 3,724 2,606No-till n.a. 1,774 1,819No-till + R 832 2,291 2,070No-till + F 2,675 3,104 2,150No-till + R + F 2,161 4,901 2,654No-till + R + F + H 2,314 5,180 3,269
Chitedze, Malawi Conventional tillage 2,420 2,522Conventional tillage + F 4,199 4,990No-till 1,667 1,942No-till + R 2,379 2,648No-till + F 3,372 4,828No-till + R + F 4,314 5,079No-till + R + F + H 4,321 5,524
Barue, Mozambique Conventional tillage 1,001 2,937 2,630Conventional tillage + F 1,247 4,099 4,057No-till n.a. 1,974 1,272No-till + R 894 2,283 1,774No-till + F 1,153 3,411 2,910No-till + R + F 1,409 3,328 4,168No-till + R + F + H 1,029 3,808 4,057
Hwedza, Zimbabwe Conventional tillage 866 370Conventional tillage + F 1,769 879No-till 867 307No-till + R 1,157 355No-till + F 1,930 681No-till + R + F 2,166 975No-till + R + F + H 2,486 663
Conventional tillage + F 2,957 n.a.No-till 1,223 n.a.No-till + R 1,543 n.a.No-till + F 2,509 n.a.No-till + R + F 2,699 n.a.No-till + R + F + H 3,114 n.a.
Monze, Zambia Conventional tillage 3,577Conventional tillage + F 5,843No-till 3,087No-till + R 3,002No-till + F 5,842No-till + R + F 5,904No-till + R + F + H 4,828
SED 113 272 415
Note: Means per site in each country are based on five locations in Balaka and Hwedza, and four locations in Barue andMurehwa; at each location, the experiment is replicated four times. ∗∗Sites in Murehwa did not harvest anything due to aprolonged dry spell. R ¼ residues; F ¼ fertiliser; H ¼ herbicides.
International Journal of Agricultural Sustainability 7
Dow
nloa
ded
by [
Chr
istia
n T
hier
feld
er]
at 2
1:48
25
Janu
ary
2013
relatively more grain yield compared with no-tillage (Figure 3a). There was a positive yieldresponse to no-tillage with residues instead of without (Figure 3b). Results showed that fertiliserapplication was important as its application more than doubled yield (Figure 3c). The use of her-bicides was not beneficial over hand weeding on crop yield (Figure 3d).
Although the difference between NT with herbicides and NT with manual weeding was small,NT with herbicides was the best performing in four sites (site × season). The NT with residuesand manual weeding was the best performing treatment in only one site, i.e. Monze.
3.2. Effect of the components tested
Treatment (tillage × fertiliser application), site and season had a significant effect on maize grainyield (p , 0.001). However, the interaction between these factors was weak. Decision trees
Figure 3. Effect of key components of conservation agriculture systems on crop yield. Yields of the specificcomponent on the x-axis are put in context with yields of the other component on the y-axis. If both com-ponents are equal, they should be on the 1:1 line: (a) effect of tillage; (b) effect of residues; (c) effect of fer-tiliser in no-till systems; (d) effect of herbicides.
8 C. Thierfelder et al.
Dow
nloa
ded
by [
Chr
istia
n T
hier
feld
er]
at 2
1:48
25
Janu
ary
2013
(Figure 4) constructed through the binary recursive procedure split the data at first based on tillageand fertiliser treatment suggesting that ‘treatment’ had an overruling effect on yield. The secondmost important factor was the site; sites were split into two, based on yield potential. Monze andChitedze sites were considered as generally high-yielding sites; Balaka, Barue, Hwedza andMurehwa were grouped into the low-yielding category. Hwedza could be further isolated intoits one group as the site with the least yield potential. The least important factor was the croppingseason; the 2009 season was considered the lowest yielding compared with seasons after 2009.The effect of season on crop yield was apparent but the differences between treatments andcontrol were maintained over the experimental period.
Overall, fertiliser application had a significant effect on crop yield regardless of tillage(Figures 3c and 4). Conventional tillage with sufficient fertiliser application yielded just asmuch as no-till with fertiliser application (Table 3). The omission of CA components had a sig-nificant effect on crop yield. Generally, there was more treatment effects in Malawi, Mozambiqueand Zambia compared with the Zimbabwean sites.
4. Discussion
The field results suggest that nutrient management and not the tillage has an overriding effect oncrop yield despite the diversity in biophysical conditions. The retention of crop residues in ade-quate quantities to provide soil cover and fertiliser application under no-till practices provides themost yield benefits for farmers (Figure 3b). The sites studied often suffered mid-season dry spells;thus retention of mulch had an important role in moisture conservation and subsequent yieldbenefits compared with conventional tillage. However, this benefit only became apparent whenresidue application was combined with fertiliser. A number of authors have reported mulch reten-tion as very important in semi-arid and sub-humid environments despite the challenge to findenough biomass for residue retention (Lal 1978, Roth et al. 1988, Mupangwa et al. 2007, Wall2007, Thierfelder and Wall 2009). Besides increased crop yields, no-tillage with adequate cropresidue retention has been found to maintain higher levels of soil organic matter, total N, andexchangeable bases than tilled plots (Lal 1976). These are likely to contribute to improvementin soil fertility and larger yields in the long term (Govaerts et al. 2005).
No-till practices without mulch depressed yields compared to conventional tillage. No-tillwithout mulch cover often leads to soil crusting, reduce infiltration and moisture available tothe crops (Thierfelder et al. 2005). When rainfall is sufficient and well distributed, the effect oftillage method on plant growth, root distribution, and crop yield is often minimized comparedwith drier years (Cassel et al. 1995). Intense rainfall on the unprotected soil surface will leadto surface sealing which may reduce yields (Freese et al. 1993). These findings suggest that ifinsufficient soil cover is available, routine and systematic disturbance of the soil surface withhand-hoes might be a useful strategy for breaking up soil surface crusts which could temporarilyincrease water infiltration (Belnap 1995). However, this is not a sustainable practice as this willfurther break down organic matter through increased aeration and higher decomposition (Thier-felder and Wall 2012).
Response to fertiliser, especially N, was very high; this confirms that N is the most limitingnutrient in the sites studied. Relatively more N input is required under NT than under CT althoughthe yield differences between tillage methods are often small under semi-arid conditions (Liu andWiatrak 2012). Smallholder farmers need to invest in fertilisers in order to realise the full benefitsassociated with CA. However, the results show that fertiliser is equally required for conventionalagriculture to reap the benefits.
The lack of a clear response to herbicides application in comparison to hand-weeding suggeststhat hand-weeding is equally effective in controlling weeds or that herbicides application is as
International Journal of Agricultural Sustainability 9
Figure 4. Decision trees constructed by the binary recursive partitioning procedure for maize grain yield (2008–2011) from ‘step trials’ in southern Africa. For eachnode, the average yield and the number of cases (n) are given. Treatments are described in Table 1.
10C
.T
hierfelderet
al.
Dow
nloa
ded
by [
Chr
istia
n T
hier
feld
er]
at 2
1:48
25
Janu
ary
2013
effective as hand-weeding. Herbicides in combination with minimal weeding have been found tobe effective and economically beneficial to farmers in no-tillage systems (Mishra and Singh 2012,Ngwira et al. 2012, Rashid et al. 2012). However, in most smallholder farming systems, theabsence of opportunity costs for labour could make herbicides more expensive in the perceptionof farmers. The weeding with hand-hoes results in increased soil movement, especially whencomplicated weed species such as couch grass (Cynodon dactlon L.) or wandering jew (Comme-lina benghalensis L.) are present and farmers heavily disturb the soil to eliminate the stolon rootsof these weed species. Herbicides should therefore be used in such cases to ensure minimum soilmovement.
Although it is very difficult to assign a specific effect to the individual components of the CAsystem due to its mutually reinforcing characteristics, fully factorial trials on CA that separate theeffect of each component from others (Rusinamhodzi et al. 2011) may provide useful insights intothe starting point to integrate CA into the smallholder farming systems. It will, however, benecessary to also incorporate the principle of rotation and/or association to get a clear answerto this research question.
A step-wise incorporation of CA principles into the farming system could be one strategy thatmay allow farmers to practice CA without making drastic changes to their present farming system(Figure 5). Farmers may start by not tilling the soil and gradually concentrate available resourceson their fields, increase the productivity, and thereby increase the biomass production that mayprovide both livestock feed and soil cover. High biomass production has been found as one ofthe key requirements for the successful practice of CA in mixed crop–livestock systems (Val-buena et al. 2012). However, as our results show, no-tillage without residues and fertiliser appli-cation may lead to yield penalties in the beginning. Reduction in risk and the ability to meet foodsecurity as well as feed and mulch requirements will facilitate the widespread uptake of the tech-nology over time.
The results show that the largest yield in CA systems can be expected if no-tillage, residue reten-tion and fertilisation are implemented from the beginning. A slightly different approach could there-fore be more beneficial for smallholder farmers when sufficient resources are available. Farmerscould, instead of a step-wise implementation of one component after the other, simultaneouslyapply all principles on small areas of their farms. By concentrating their inputs (fertiliser andmaybe herbicides) and resources (residues) on a small manageable area they can increase the pro-ductivity of this land much faster and benefit more in the short term. Once they have mastered themanagement and other resource demands on this small piece of land, they can expand to other areasdepending on their landholding, and increase yield, income and food security over time.
Table 3. Crop yield as affected by tillage, mulch, fertiliser and weed control across the experimental sites.
Treatment
MalawiMozambique
ZimbabweZambia
Balaka Chitedze Barue Hwedza Murehwa Monze
CT 1,593 2,471 2,425 618 1,463 3,577CT + F 2,833 4,595 3,392 1,324 2,957 5,843No-till 1,796 1,805 1,573 587 1,223 3,087No-till + R 1,893 2,514 1,748 756 1,543 3,002No-till + F 2,586 4,100 2,687 1,306 2,509 5,842No-till + R + F 3,188 4,697 3,275 1,570 2,700 5,904No-till + R + F + H 3,526 4,923 3,301 1,574 3,114 4,828SED 254 494 263 150 375 715
R, residues; F, fertiliser; H, herbicides.
International Journal of Agricultural Sustainability 11
Dow
nloa
ded
by [
Chr
istia
n T
hier
feld
er]
at 2
1:48
25
Janu
ary
2013
5. Conclusion
Retention of previous crop harvest residues and fertiliser application is critical for the success ofCA systems. When insufficient crop residues are available for soil cover, soil crusting and sealingmay occur, leading to reductions in rainfall infiltration and ultimately small crop yields. Resultssuggest that N deficiencies should be addressed in order to realise the full benefits of CA systems.Hand-weeding and herbicides application are both effective in weed control; herbicides useshould, however, be promoted in order to reduce soil movement. A step-wise integration ofCA in the smallholder farming systems is proposed as a possible strategy to avoid creating con-straints. However, if sufficient resources including adequate amounts of crop residues are avail-able, practising all principles of CA from the onset may provide the best yield in the croppingsystem.
AcknowledgementsWe are grateful to Mphatso Gama and Ivy Ligowe (Malawi), Sign Phiri and Herbert Chipara(Zimbabwe), Felisardo Chicoche and Zacarias Bango (Mozambique) and Mwangala Sitali (Zambia)for the management of field experiments and data collection. Financial support from CIMMYT andCIAT through the SSA-CP project funded by FARA and the CA project funded by IFAD is sincerelyappreciated.
Figure 5. A sustainable pathway to the full integration of CA in the smallholder farming systems ofsouthern Africa.
12 C. Thierfelder et al.
Dow
nloa
ded
by [
Chr
istia
n T
hier
feld
er]
at 2
1:48
25
Janu
ary
2013
ReferencesApte, C. and Weiss, S., 1997. Data mining with decision trees and decision rules. Future generation com-
puter system, 13, 197–210.Bayer, C., et al., 2000. Organic matter storage in a sandy clay loam Acrisol affected by tillage and cropping
systems in southern Brazil. Soil and tillage research, 54, 101–109.Belnap, J., 1995. Surface disturbances: Their role in accelerating desertification. Environmental monitoring
and assessment, 37, 39–57.Bessam, F. and Mrabet, R., 2003. Long-term changes in soil organic matter under conventional tillage and
no-tillage systems in semiarid Morocco. Soil use manage, 19, 139–143.Bolliger, A., et al., 2006. Taking stock of the Brazilian ‘zero-till revolution’: a review of landmark research
an farmers’ practice. Advances in agronomy, 91, 47–110.Buntine, W., 1992. Learning classification trees. Statistics and computing, 2, 63–73.Cassel, D.K., Raczkowski, C.W., and Denton, H.P., 1995. Tillage effects on corn production and soil phys-
ical conditions. Soil science society of America Journal, 59, 1436–1443.Dixon, J., et al., 2001. Farming systems and poverty: improving farmers’ livelihoods in a changing world.
Rome and Washington DC: FAO and World Bank.Erenstein, O., 2002. Crop residue mulching in tropical and semi-tropical countries: An evaluation of residue
availability and other technological implications. Soil and tillage research, 67, 115–133.Erenstein, O., et al., 2012. Conservation agriculture in maize- and wheat-based systems in the (Sub)tropics:
Lessons from Adaptation Initiatives in South Asia, Mexico, and Southern Africa. Journal of sustainableagriculture, 36, 180–206.
FAO, 2002. Conservation agriculture: case studies in Latin America and Africa. Rome: FAO Soils Bulletin78 FAO.
Freese, R.C., Cassel, O.K., and Denton, H.P., 1993. Infiltration in a Piedmont soil under three tillage treat-ments. Journal of soil water conservation, 48, 214–218.
Gilbert, N., 2012. Dirt poor: The key to tackling hunger in Africa is enriching its soil. The big debate is abouthow to do it. Nature, 483, 525–527.
Giller, K.E., et al., 2009. Conservation agriculture and smallholder farming in Africa: The heretic’s view.Field crops research, 114, 23–34.
Govaerts, B., Sayre, K.D., and Deckers, J., 2005. Stable high yields with zero tillage and permanent bedplanting? Field crops research, 94, 33–42.
Govaerts, B., et al., 2006. Long-term consequences of tillage, residue management, and crop rotation onmaize/wheat root rot and nematode populations in subtropical highlands. Applied soil ecology, 32,305–315.
Ito, M., Matsumoto, T., and Quinones, M.A., 2007. Conservation tillage practices in sub-Saharan Africa: Theexperience of Sassaka Global 2000. Crop protection, 26, 417–423.
Kassam, A., et al., 2009. The spread of conservation agriculture: justification, sustainability and uptake.International journal of agricultural sustainability, 7, 292–320.
Lal, R., 1976. No-tillage effect on soil properties under different crops in western Nigeria. Soil sciencesociety of America journal, 40, 762–768.
Lal, R., 1978. Importance of tillage system in soil and water management in the tropics, soil tillage and CropProduction. Ibadan, Nigeria: IITA, 25–32.
Lal, R., Kimble, J.M., and Stewart, B.A., 2000. Global climate change and cold regions ecosystems. BocaRaton, FL: Lewis Publishers.
Liu, K. and Wiatrak, P., 2012. Corn production response to tillage and nitrogen application in dry-landenvironment. Soil and tillage research, 124, 138–143.
Mafongoya, P.L., et al., 2006. Appropriate technologies to replenish soil fertility in southern Africa. Nutrientcycling in agroforestry, 76, 137–171.
Mishra, J.S. and Singh, V.P., 2012. Tillage and weed control effects on productivity of a dry seeded rice–wheat system on a Vertisol in Central India. Soil tillage research, 123, 11–20.
Mueller, J.P., et al., 2001. Conflicts between conservation agriculture and livestock over utilization of cropresidues. In: J. Garcia-Torres, A. Benites and Mart¡nez-Vilela, eds. Conservation agriculture: a world-wide challenge. C¢rdoba, Spain: ECAF/FAO, 211–225.
Mupangwa, W., et al., 2007. Effect of minimum tillage and mulching on maize (Zea mays L.) yield and watercontent of clayey and sandy soils. Physics and chemistry of the earth, parts A/B/C, 32, 1127–1134.
Ngwira, A.R., Aune, J.B., and Mkwinda, S., 2012. On-farm evaluation of yield and economic benefit of shortterm maize legume intercropping systems under conservation agriculture in Malawi. Field cropsresearch, 132, 149–157.
International Journal of Agricultural Sustainability 13
Rashid, M.H., et al., 2012. Comparative efficacy of pretilachlor and hand weeding in managing weeds andimproving the productivity and net income of wet-seeded rice in Bangladesh. Field crops research, 128,17–26.
Roth, C.H., et al., 1988. Effect of mulch rates and tillage systems on infiltrability and other soil physicalproperties of an Oxisol in Parafla, Brazil. Soil and tillage research, 11, 81–91.
Rusinamhodzi, L., et al., 2011. A meta-analysis of long-term effects of conservation agriculture on maizegrain yield under rain-fed conditions. Agronomy for sustainable development. 1–17.
Sanchez, P., 2002. Soil fertility and hunger in Africa. Science, 295, 2019–2020.Smaling, E.M.A., Nandwa, S.M., and Janssen, B.H., 1997. Soil fertility in Africa is at stake. In: R.J. Buresh
and P.A. Sanchez, eds. Replenishing soil fertility in Africa. SSSA Special Publication SSSA, Madison,WI, USA, 47–61.
Statistix, 2008. Statistix 9. Tallahassee, FL: Analytical Software.Thierfelder, C., Amezquita, E., and Stahr, K., 2005. Effects of intensifying organic manuring and tillage
practices on penetration resistance and infiltration rate. Soil and tillage research, 82, 211–226.Thierfelder, C. and Wall, P.C., 2009. Effects of conservation agriculture techniques on infiltration and soil
water content in Zambia and Zimbabwe. Soil and tillage research, 105, 217–227.Thierfelder, C. and Wall, P.C., 2011. Reducing the risk of crop failure for smallholder farmers in Africa
through the adoption of conservation agriculture. In: A. Bationo, B. Waswa, J.M.M. Okeyo, F.Maina, and J.M. Kihara, eds. Innovations as Key to the Green Revolution in Africa. Netherlands:Springer, 1269–1277.
Thierfelder, C. and Wall, P.C., 2012. Effects of conservation agriculture on soil quality and productivity incontrasting agro-ecological environments of Zimbabwe. Soil use and management, 28, 209–220.
Tittonell, P., et al., 2007. Nutrient use efficiencies and crop responses to N, P and manure applications inZimbabwean soils: Exploring management strategies across soil fertility gradients. Field crops research,100, 348–368.
Valbuena, D., et al., 2012. Conservation Agriculture in mixed crop–livestock systems: Scoping crop residuetrade-offs in Sub-Saharan Africa and South Asia. Field crops research, 132, 175–184.
Vanlauwe, B., et al., 2010. Integrated soil fertility management: Operational definition and consequences forimplementation and dissemination. Outlook on agriculture, 39, 17–24.
Verhulst, N., et al., 2010. Conservation agriculture, improving soil quality for sustainable productionsystems. In: R. Lal and B.A. Stewart, eds. Advances in soil science: food security and soil quality.Boca Raton, FL: CRC Press, 137–208.
Wall, P.C., 1999. Experiences with crop residue cover and direct seeding in the Bolivian highlands. Mountainresearch and development, 19, 313–317.
Wall, P.C., 2007. Tailoring conservation agriculture to the needs of small farmers in developing countries: ananalysis of issues. Journal of crop improvement, 19, 137–155.
World Bank, 2007. International development association on a country assistance strategy of the WorldBank for the Republic of Malawi. Washington, DC: World Bank.
